AI adoption will slow office-based employment growth in New Zealand but will not reverse it, with job numbers set to keep rising and office space demand increasing in tandem, according to new analysis from CBRE Research.
The ‘Investigations of potential AI impacts on office employment and demand’ study, conducted over the three months to June 2026, modelled AI adoption scenarios across two contrasting Auckland office precincts: the CBD and a suburban commercial centre.
Office-based employment kept growing in every scenario, at up to 0.9% a year, compared with around 1.3% growth without AI.
Zoltan Moricz, head of research at CBRE New Zealand, said the research shows the doom and gloom scenarios many people feared about AI’s impact on office-based jobs are likely to be overplayed.
Start your property search
“We adopted in-depth labour market analytics on the impact of AI adoption and overlaid it with our own analysis of how this could impact office space demand. Our results show that there is still going to be office employment growth, although not to the same extent as in a world without AI,” he said.
“While the rate of growth will be somewhat lower as a result of AI adoption, we’re unlikely to see wide-ranging automation of roles, or any significant contractionary impact on employment and office space demand.”
One of the study’s key findings is that AI is far more likely to augment office work - enhancing productivity and efficiency - than fully automate it.
In addition, enhanced productivity with AI use should translate to profitability gains which will support the prime grade office leasing market by improving rental affordability, Moricz said.
“The flight to quality among office occupiers continues, but it's not just about wanting better space - you need to be able to afford the rent. We believe AI will have a positive impact on business’ bottom lines through enhancing productivity in office-based occupations.”
Analysis of 2.7m jobs across New Zealand found just 4% are highly exposed to AI automation, while 32% are highly exposed to augmentation.
Occupational exposure figures also frequently overstate what AI can realistically do. A paper on the labour market impacts published earlier this year by Anthropic counts a task as exposed if AI can perform any part of it. For example, legal sector jobs were theoretically exposed to around 90% automation in Anthropic’s report.
In contrast, CBRE’s task-by-task analysis of how lawyers actually spend their time, and where AI can practically be used, puts task coverage at about 30%, said Moricz.
“AI task coverage incorporates a much more significant degree of augmentation rather than automation. To state the obvious, AI can’t do court appearances and client meetings.”
Based on Claude usage, occupational coverage appears to be plateauing, at least for the time being. The percentage of occupations (across all categories, not just office-based) where at least 25% of tasks are exposed to Claude is plateauing at just below 50%, according to Anthropic.
Occupations with higher AI penetration (50+ per cent and 75+ per cent of task coverage) are growing at a gradual level, at similar rates to those seen during the earlier stages of adoption. As at February 2026, AI use covers 75% or more of tasks for 7% of occupations and 50% or more of tasks for 24% of occupations.
New Zealand, as an early adopter of AI, is well placed to model the evolution of these trends - ranking 13th out of 116 countries for depth of usage of Claude.
Effects will differ between office markets depending on their occupational mix. Centres weighted towards back office and support functions face heavier impacts on employment growth and office space demand, with administrative and sales roles more exposed to automation.
In contrast, markets dominated by head offices, client-facing roles and professional occupations (such as CBDs) will see softer effects. CBRE's modelling puts the difference at around 0.3 percentage points of employment growth a year.
Slower employment growth will also have a lesser impact on office demand than it would have a decade ago. CBRE’s analysis shows the long-running disconnect between white collar employment growth and office take-up is stabilising, after the two moved closely together until around 2010 and then diverged sharply.
“Even with a reduced rate of employment growth because of AI, more of that employment growth is likely to flow through to the office market as demand for space,” Moricz said. “This premise should be central to the discussion as we move into a maturing usage phase with AI.”
- Supplied by CBRE
